personaforgeAI agents that ship.

One framework. From hello-world to hardened enterprise — durability, guardrails, budgets, and audit included. In one package. With zero lock-in.

Get Started → 5 minutes

See Examples

Trust & Reliability →

100+Built-in Tools

20+Vector Stores

30+LLM Providers

MITOpen Source

SEE IT IN ACTION

From idea to production, in minutes

Every pattern you'll ever need — from a one-liner to a hardened enterprise agent.

hello.tsTypeScript

import { agent } from 'personaforge';

// That's it. Model, session & guardrails wired automatically.
const ai = agent('You are a helpful assistant.');

const { text } = await ai.run(
  'Summarize the Rust ownership model in 3 bullets.',
);

console.log(text);

Why personaforge?

Most agent frameworks give you an LLM wrapper and call it a day. personaforge is different — it ships every production primitive in a single npm install. No stitching together 7 packages. No "choose your own adventure" with sessions, memory, and guardrails.

Capability	personaforge	LangChain.js	Vercel AI SDK	CrewAI	LangGraph	Mastra	Agno
Zero-config progressive DX	✓	~	✓	~	~	~	~
First-class TypeScript	✓	~	✓	✗	~	✓	✗
100+ built-in tools	✓	✓	✗	~	✗	~	~
Multi-agent orchestration	✓	✓	✗	✓	✓	✓	✓
Durable DAG graph engine	✓	~	✗	✗	✓	✗	~
Native MCP support	✓	~	✗	~	~	✓	~
OTLP distributed tracing	✓	~	~	✗	~	~	✗
Circuit breakers & retries	✓	✗	✗	✗	~	✗	✗
USD budget enforcement	✓	✗	✗	✗	✗	✗	✗
Multi-tenancy context	✓	✗	✗	✗	✗	✗	✗
Audit logging	✓	✗	✗	✗	✗	✗	✗
Human-in-the-loop (HITL)	✓	~	✗	~	✓	~	~
Intelligent LLM router	✓	✗	✗	✗	✗	✗	~
Automatic REST API	✓	✗	✗	✗	✗	~	~
Voice & video	✓	~	✗	✗	✗	✗	✗

One agent, one run — start here

import { agent } from 'personaforge';

const bot = agent('You are a helpful assistant.');

const { text } = await bot.run('What is the capital of France?');
console.log(text); // "The capital of France is Paris."

Zero config. The framework resolves the provider from your environment, wires sessions in memory, and runs. No boilerplate. No ceremony.

Add capabilities as you need them

import { agent, tool, InMemorySessionStore, Memory } from 'personaforge';
import { z } from 'zod';

const searchWeb = tool({
  name: 'search_web',
  description: 'Search the web.',
  parameters: z.object({ query: z.string() }),
  execute: async ({ query }) => fetch(`https://api.search.com?q=${query}`).then(r => r.json()),
});

const memory = new Memory({ options: { lastMessages: 20 } });

const researchAgent = agent({
  name: 'researcher',
  model: 'gpt-4o',
  instructions: 'Research topics using web search. Remember preferences.',
  tools: [searchWeb],
  memory,
  sessionStore: new InMemorySessionStore(),
  dev: true, // console logging when you're iterating
});

Scale to production — same API

import { createEnterpriseGateway, createAgent } from 'personaforge';

const support = createAgent({ name: 'support', model: 'gpt-4o', instructions: '...' });
const billing = createAgent({ name: 'billing', model: 'gpt-4o', instructions: '...' });

const gateway = createEnterpriseGateway({
  agents: { support, billing },
  auth: apiKeyAuth([process.env.GATEWAY_API_KEY!]),
  tenants: [
    { id: 'acme', monthlyBudgetUsd: 500, maxRpm: 60, allowedAgents: ['support'] },
  ],
  policy: { monthlyBudgetUsd: 5000 },
});

await gateway.start(8787);
// → SOC 2 / HIPAA / GDPR dashboard at /compliance

Three primitives. Every pattern.

Primitive	When
Agent	One model-backed worker with tools, memory, and state
Team	Specialists coordinating — supervisor, consensus, swarm, handoff
Workflow	Staged execution with branching, parallelism, and durability

BATTERIES INCLUDED

100+ tools, zero assembly required

Every tool is Zod-validated, tree-shakeable, and ready to drop into any agent. Import only what you need — each subpath is independently bundleable.

🌐HTTP & Web

fetchscrapebrowsercrawlsitemap

💬Communication

emailSlackDiscordTwilioTelegram

🗄️Databases

PostgreSQLMySQLSQLiteRedisMongoDB

☁️Cloud Storage

S3GCSAzure BlobDropboxDrive

🔧Dev Tools

GitHubGitLabJiraLinearNotion

🔍Search & Data

DuckDuckGoWikipediaCSVJSONExcel

💳Payments

StripePayPalinvoicessubscriptionsrefunds

📁File System

readwritecopymovezip/unzip

🧮Compute & Math

calculatorcode runnerunit convertdate/timecron

Tree-shakeable subpath importsImport only what you need

personaforgeMain barrel export

personaforge/tools100+ built-in tools

personaforge/orchestrationMulti-agent patterns

personaforge/knowledgeRAG + vector store

personaforge/sessionSession stores

personaforge/guardrailsSafety & validation

personaforge/productionCircuit breakers, rate limits

personaforge/observabilityOTLP, logs, evals

personaforge/runtimeHTTP + WebSocket server

personaforge/adapters20-category adapters

personaforge/testingMocks & fixtures

personaforge/contractsShared types only

LLM PROVIDERS

One framework, every model

30+ providers built-in. Swap with one import — the same agent code runs on every model.

OpenAIGPT & o-series

OPENAI_API_KEY

AnthropicClaude family

ANTHROPIC_API_KEY

GoogleGemini family

GOOGLE_API_KEY

OpenRouter100+ models

OPENROUTER_API_KEY

Azure OpenAIEnterprise

OPENAI_BASE_URL

AWS BedrockPeer dep

@aws-sdk/...

Any OpenAI-compatCustom URL

apiKey + baseURL

OllamaLocal

No key needed

Built for production — by default

Capability	Ships in the box
Guardrails	PII detection, prompt injection defense, content moderation, tool/host allowlists
Budget enforcement	Per-user, per-session, and global USD caps — stop before the bill surprises you
Rate limiting	Sliding-window (in-memory + Redis)
Circuit breakers	Automatic provider failure detection and recovery
HITL approvals	Human-in-the-loop tool approval with `interrupt()` / `resume()`
Audit trail	Hash-chained event log, SOC 2 ready
OTLP tracing	OpenTelemetry with gen-ai semantic conventions
Prometheus metrics	Request counts, latency, token usage, error rates out of the box
Durable execution	Checkpoint/restore, deterministic replay from event log
Multi-tenancy	Per-tenant budgets, rate limits, agent allowlists, RBAC
Graceful shutdown	Drain active executions before shutdown
Health checks	Readiness, liveness, dependency probes

Eval built in

import { evaluate, fromAgent } from 'personaforge';

const results = await evaluate({
  subject: fromAgent(myAgent),
  dataset: [
    { input: 'What is 2+2?', expected: { contains: '4' } },
    { input: 'Capital of Japan?', expected: { contains: 'Tokyo' } },
  ],
  metrics: ['accuracy', 'latency', 'cost'],
  judge: { model: 'gpt-4o', criteria: ['correctness', 'conciseness'] },
});

console.log(results.passRate); // 0.92

Framework comparisons

COMPETITIVE LANDSCAPE

Enterprise-ready from day one

No patchwork of add-ons. Every production concern is built-in and composable.

Capability	personaforge	LangChain.js	Vercel AI SDK	CrewAI	LangGraph	Mastra	Agno
Zero-config progressive DX	✓	~	✓	~	~	~	~
First-class TypeScript	✓	~	✓	✗	~	✓	✗
100+ built-in tools	✓	✓	✗	~	✗	~	~
Multi-agent orchestration	✓	✓	✗	✓	✓	✓	✓
Durable DAG graph engine	✓	~	✗	✗	✓	✗	~
Native MCP support	✓	~	✗	~	~	✓	~
OTLP distributed tracing	✓	~	~	✗	~	~	✗
Circuit breakers & retries	✓	✗	✗	✗	~	✗	✗
USD budget enforcement	✓	✗	✗	✗	✗	✗	✗
Multi-tenancy context	✓	✗	✗	✗	✗	✗	✗
Audit logging	✓	✗	✗	✗	✗	✗	✗
Human-in-the-loop (HITL)	✓	~	✗	~	✓	~	~
Intelligent LLM router	✓	✗	✗	✗	✗	✗	~
Automatic REST API	✓	✗	✗	✗	✗	~	~
Voice & video	✓	~	✗	✗	✗	✗	✗

Migration guides for LangChain, CrewAI, LangGraph, Mastra & Agno →

WHY TRUST PERSONAFORGE

Built for production, proven in CI

Capability claims are easy. We publish the evidence — tests, benchmarks, security policy, and governance.

1,500+Automated TestsHermetic MockLLM — no live API calls in CI

92%τ-bench Pass Rate12/13 tool-calling tasks on gpt-4o-mini

43%+CI Coverage FloorRatcheting quarterly toward 75%

MITOpen SourceNo telemetry by default · zero lock-in

🔒Security-first

Private vulnerability disclosure, JWT hardening, SSRF-protected tools, and a production hardening checklist.

Learn more →

🧪Tested in CI

1,500+ tests across 124 files. MockLLMProvider for deterministic runs. Coverage gates enforced on every PR.

Learn more →

📊Benchmarked

τ-bench harness with verifier-based scoring. Cross-framework protocol for LangGraph, Agno, CrewAI, and Mastra.

Learn more →

🔭Observable

OpenTelemetry-native tracing, tamper-evident audit logs, Prometheus metrics, and a built-in control plane.

Learn more →

Read the full trust & reliability guide →

DELIGHTFUL DX

Built for developer happiness at every scale

Zero-to-agent in 3 lines. A path to enterprise that never forces a rewrite. Every abstraction earns its place.

Install one package

npm install personaforge

No 12-step setup. No mandatory config files.

Set an API key

OPENAI_API_KEY=sk-...

Any of 30+ providers. Swap at any time.

Run your first agent

import { agent } from 'personaforge';
const { text } = await agent('Be helpful.').run('Hello!');

Smart defaults chosen for you. Override anything.

🧩

Progressive Escape Hatches

Start with agent() one-liner. Add tools, sessions, guardrails, budgets one-by-one as you need them. Never rewrite.

📐

Full TypeScript Inference

Every parameter, every hook, every tool result is typed end-to-end. Autocomplete works everywhere — no any.

🧪

Test Without an LLM

MockLLMProvider + MockToolRegistry let you write fast, deterministic unit tests without real API calls.

⚡

Smart Defaults, Not Magic

Defaults are explicit and documented. No hidden global state. No invisible retry loops. Every behaviour is opt-in.

🔀

Mix and Match

Combine createAgent(), compose(), createSupervisor() freely. The abstractions are composable, not hierarchical.

📦

Monorepo Friendly

Independent subpath imports mean each service only bundles what it needs. Works perfectly with Turborepo, Nx, and Bun workspaces.

ENTERPRISE CHECKLIST

Everything you need to go to production

No switching frameworks at scale. Every capability ships with personaforge.

🔒Security

Guardrails engine with sensitive-data rules
JWT RBAC on HTTP routes
Secret-manager adapters (AWS, Azure KV, HashiCorp, GCP)
Content safety hooks

⚡Reliability

Circuit breakers with configurable thresholds
Exponential-backoff retry with jitter
Redis distributed rate limiting
Graceful shutdown + checkpoint/resume

📋Compliance

Persistent audit log (SQLite / pluggable)
X-Idempotency-Key deduplication
Per-user and per-tenant cost caps
W3C trace-context propagation

🔭Observability

OTLP tracing (Jaeger, Datadog, Honeycomb)
Structured logging with context
Eval store + LLM-as-judge scoring
Health endpoints + Grafana dashboard

🚀Deployment

Docker + docker-compose templates
Kubernetes with rolling updates & probes
Fly.io and Render one-click config
OpenAPI + WebSocket + SSE server built-in

🧪Testing

MockLLMProvider for deterministic tests
MockToolRegistry + fixture helpers
Vitest-compatible test utilities
Deterministic tests — no live API calls

READY TO SHIP?

Start building in
30 seconds

One package. From prototype to enterprise without changing frameworks.

npm install personaforge

Get Started → View on GitHub

MIT License·No telemetry by default·TypeScript-first·Zero lock-in

personaforgeAI agents that ship.

From idea to production, in minutes

Why personaforge? ​

One agent, one run — start here ​

Add capabilities as you need them ​

Scale to production — same API ​

Three primitives. Every pattern. ​

100+ tools, zero assembly required

One framework, every model

Built for production — by default ​

Eval built in ​

Framework comparisons ​

Enterprise-ready from day one

Built for production, proven in CI

Built for developer happiness at every scale

Everything you need to go to production

Start building in30 seconds

Why personaforge?

One agent, one run — start here

Add capabilities as you need them

Scale to production — same API

Three primitives. Every pattern.

Built for production — by default

Eval built in

Framework comparisons

Start building in
30 seconds